Traffic Jam in Special Capital Region (DKI) Jakarta has become a problem since the beginning until now that bothers people to run their activities. To minimize this problem, people are looking for, and exchange information about congestion on social media. One of the social media that is frequently used is by people of the Capital City is Twitter.Twitter is one of the social media in Indonesia that has usesrs increased from year to year . Modern society people like Jakarta is also utilizing social media to exchange information, including tarffic jam. One of a Twitter account that provides information about traffic is @TMCPoldaMetro Twitter account, official accounts belonging Jakarta’s Police City . Every day, an average of 200-300 informs this account tweets about congestion in Jakarta from morning till night. With the huge amount of data, then can be analyzed to determine the pattern Datamining of a certain data.
In this study, We created a classification system congestion in Jakarta with one of the data mining techniques, namely classification. By using Naive Bayes classification method, the data are taken from congestion Twitter accountTMCPoldaMetro will be processed and used as a basis for predicting the possible next traffic jam that will happen.Then, from the results of performance testing for five times with different training data in each test, each obtained -accuracy results, the first test 56.25%, the second test 63%; The third test 58.89%; The fourth test 61.54%, the highest accuracy of testing with the test contained in the fifth test with 68.66% accuracy.